Summary Viral evolution enables the emergence of novel viral pathogens in the human population. The evolution of influenza A virus is also critical for the maintenance of seasonal lineages in the context of host immunity. Evolution can result from selection, leading to increased fitness, or stochastic processes that typically decrease fitness. The relative potency of selective and stochastic forces is therefore a critical determinant of the adaptive potential of a population. Based on evolutionary theory, we hypothesize that the magnitude of stochastic effects in viral evolution is strongly impacted by the spatial structure that characterizes viral spread within a host. In other words, the expansion of a virus population in space may give rise to random, within-host bottlenecks and founder effects that weaken the efficiency of natural selection. Factors that shape viral spread - including viral phenotypes, host responses and physical characteristics of the host environment - are therefore predicted to impact viral genetic diversity and evolution. Our overarching hypothesis is that viral features that modulate viral spatial structure also modulate viral diversity and evolution. We will test this hypothesis for influenza A virus using a well-integrated combination of simulation modeling and experimental approaches. In Aim 1, we will use computational, cell culture and ferret models to examine the consequences of spatially structured spread for genetic diversity of viral populations. For our experiments, we will use viruses carrying a selectively neutral barcode to allow robust quantification of viral diversity. In this way, the degree to which stochastic effects dominate viral dynamics will be examined under a range of conditions. In Aim 2, we will evaluate the consequences of spatially structured spread for both purifying and positive selection. Computational approaches will be used to develop hypotheses of how long distance virus dispersal impacts the ability of de novo beneficial and deleterious mutations to reach dominance. These model predictions will then be tested using experimental evolution under conditions of common vs. rare long distance dispersal. In this way, we will test the theoretical concept that stochastic effects associated with spatial structure impede the ability of natural selection to act on expanding populations. Taken together, the research proposed in these two aims will uncover the importance of spatial structure to viral population biology and evolution, deepening our fundamental understanding of the forces shaping viral evolution in nature.